Most mentoring platforms fall into one of two traps:
- Pure directory / swipe model - you’re on your own to find someone.
- “Smart” algorithmic assignment - the platform decides who you should talk to.
Both have problems.
The swipe model is exhausting
When there are hundreds of profiles, good mentors get buried. Mentees burn time on people who are a bad fit on paper. The signal-to-noise ratio is terrible.
Algorithmic assignment removes agency
Even the best matching algorithms are imperfect. They optimize for surface-level signals (skills listed, years of experience, keywords). They miss nuance, personality fit, current bandwidth, and whether someone is actually the right kind of mentor for where you are right now.
When the platform assigns you someone, two bad things happen:
- You feel less ownership over the relationship.
- If it doesn’t work, you blame the platform instead of learning how to evaluate mentors yourself.
Our approach: AI surfaces, humans decide
We use AI to do the expensive part - scanning profiles, goals, and past interactions to surface a small number of high-potential matches.
Then we stop.
You get to browse the shortlist. You get to read their writeups and backgrounds. You get to choose.
This preserves two important things:
- Agency - you own the decision.
- Learning - you get better at judging who is right for you over time.
The AI is a scout, not a matchmaker.
This is harder to build
It would be easier to just auto-book people. It would also be worse.
We believe the friction of “I have to choose” is valuable friction. It forces clarity on what you actually want.
If you’re a mentee, take the recommendations seriously - but don’t outsource the final decision.
If you’re a mentor, know that the people who reach out to you are people who actively chose you. That changes the quality of the conversation.
AI helps us see more clearly.
Humans still have to decide.
That’s the model.